Execution Flow

This document explains how CyanPrint executes templates, processors, and plugins through the Boron coordinator using containerized environments.

Overview

When a user runs a command like cyanprint create <template_ref> [path], the following high-level flow occurs:

Detailed Execution Phases

Phase 1: Template Resolution

The CLI first resolves the requested template:

During this phase:

1. The template identifier is parsed (name, version, registry)
2. Template metadata is fetched from the Zinc registry
3. Dependencies are resolved transitively
4. Input schema is retrieved for validation

Phase 2: Input Collection

Once the template is resolved, inputs are collected:

Inputs are collected through interactive prompts driven by the template's inquirer schema. The template defines the questions and validation rules, and the SDK handles user interaction.

Phase 3: Container Scheduling

Boron manages container execution:

Phase 4: Template Execution

Inside the container, the template executes:

The template uses the Helium SDK to:

1. Access validated inputs
2. Generate files and directories
3. Set file permissions and metadata
4. Return a virtual file tree

Phase 5: Processor Pipeline

Processors transform the template output in parallel:

Processors are primarily templaters that perform file transformations:

• Template engines: Handlebars, Mustache, Jinja2, etc.
• Syntax transformers: TypeScript compiler, Sass, etc.
• File processors: Copy, rename, organize files

Phase 6: Plugin Application

Plugins can modify the final output:

Phase 7: Output Streaming

Results are streamed back to the client:

Container Lifecycle

Error Handling

Next Steps

• Learn about specific Repositories
• Set up your Development Environment
• Return to Architecture Overview